1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
|
// Copyright 2012-2023 The NATS Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package nats
import (
"fmt"
)
type msgArg struct {
subject []byte
reply []byte
sid int64
hdr int
size int
}
const MAX_CONTROL_LINE_SIZE = 4096
type parseState struct {
state int
as int
drop int
hdr int
ma msgArg
argBuf []byte
msgBuf []byte
msgCopied bool
scratch [MAX_CONTROL_LINE_SIZE]byte
}
const (
OP_START = iota
OP_PLUS
OP_PLUS_O
OP_PLUS_OK
OP_MINUS
OP_MINUS_E
OP_MINUS_ER
OP_MINUS_ERR
OP_MINUS_ERR_SPC
MINUS_ERR_ARG
OP_M
OP_MS
OP_MSG
OP_MSG_SPC
MSG_ARG
MSG_PAYLOAD
MSG_END
OP_H
OP_P
OP_PI
OP_PIN
OP_PING
OP_PO
OP_PON
OP_PONG
OP_I
OP_IN
OP_INF
OP_INFO
OP_INFO_SPC
INFO_ARG
)
// parse is the fast protocol parser engine.
func (nc *Conn) parse(buf []byte) error {
var i int
var b byte
// Move to loop instead of range syntax to allow jumping of i
for i = 0; i < len(buf); i++ {
b = buf[i]
switch nc.ps.state {
case OP_START:
switch b {
case 'M', 'm':
nc.ps.state = OP_M
nc.ps.hdr = -1
nc.ps.ma.hdr = -1
case 'H', 'h':
nc.ps.state = OP_H
nc.ps.hdr = 0
nc.ps.ma.hdr = 0
case 'P', 'p':
nc.ps.state = OP_P
case '+':
nc.ps.state = OP_PLUS
case '-':
nc.ps.state = OP_MINUS
case 'I', 'i':
nc.ps.state = OP_I
default:
goto parseErr
}
case OP_H:
switch b {
case 'M', 'm':
nc.ps.state = OP_M
default:
goto parseErr
}
case OP_M:
switch b {
case 'S', 's':
nc.ps.state = OP_MS
default:
goto parseErr
}
case OP_MS:
switch b {
case 'G', 'g':
nc.ps.state = OP_MSG
default:
goto parseErr
}
case OP_MSG:
switch b {
case ' ', '\t':
nc.ps.state = OP_MSG_SPC
default:
goto parseErr
}
case OP_MSG_SPC:
switch b {
case ' ', '\t':
continue
default:
nc.ps.state = MSG_ARG
nc.ps.as = i
}
case MSG_ARG:
switch b {
case '\r':
nc.ps.drop = 1
case '\n':
var arg []byte
if nc.ps.argBuf != nil {
arg = nc.ps.argBuf
} else {
arg = buf[nc.ps.as : i-nc.ps.drop]
}
if err := nc.processMsgArgs(arg); err != nil {
return err
}
nc.ps.drop, nc.ps.as, nc.ps.state = 0, i+1, MSG_PAYLOAD
// jump ahead with the index. If this overruns
// what is left we fall out and process a split buffer.
i = nc.ps.as + nc.ps.ma.size - 1
default:
if nc.ps.argBuf != nil {
nc.ps.argBuf = append(nc.ps.argBuf, b)
}
}
case MSG_PAYLOAD:
if nc.ps.msgBuf != nil {
if len(nc.ps.msgBuf) >= nc.ps.ma.size {
nc.processMsg(nc.ps.msgBuf)
nc.ps.argBuf, nc.ps.msgBuf, nc.ps.msgCopied, nc.ps.state = nil, nil, false, MSG_END
} else {
// copy as much as we can to the buffer and skip ahead.
toCopy := nc.ps.ma.size - len(nc.ps.msgBuf)
avail := len(buf) - i
if avail < toCopy {
toCopy = avail
}
if toCopy > 0 {
start := len(nc.ps.msgBuf)
// This is needed for copy to work.
nc.ps.msgBuf = nc.ps.msgBuf[:start+toCopy]
copy(nc.ps.msgBuf[start:], buf[i:i+toCopy])
// Update our index
i = (i + toCopy) - 1
} else {
nc.ps.msgBuf = append(nc.ps.msgBuf, b)
}
}
} else if i-nc.ps.as >= nc.ps.ma.size {
nc.processMsg(buf[nc.ps.as:i])
nc.ps.argBuf, nc.ps.msgBuf, nc.ps.msgCopied, nc.ps.state = nil, nil, false, MSG_END
}
case MSG_END:
switch b {
case '\n':
nc.ps.drop, nc.ps.as, nc.ps.state = 0, i+1, OP_START
default:
continue
}
case OP_PLUS:
switch b {
case 'O', 'o':
nc.ps.state = OP_PLUS_O
default:
goto parseErr
}
case OP_PLUS_O:
switch b {
case 'K', 'k':
nc.ps.state = OP_PLUS_OK
default:
goto parseErr
}
case OP_PLUS_OK:
switch b {
case '\n':
nc.processOK()
nc.ps.drop, nc.ps.state = 0, OP_START
}
case OP_MINUS:
switch b {
case 'E', 'e':
nc.ps.state = OP_MINUS_E
default:
goto parseErr
}
case OP_MINUS_E:
switch b {
case 'R', 'r':
nc.ps.state = OP_MINUS_ER
default:
goto parseErr
}
case OP_MINUS_ER:
switch b {
case 'R', 'r':
nc.ps.state = OP_MINUS_ERR
default:
goto parseErr
}
case OP_MINUS_ERR:
switch b {
case ' ', '\t':
nc.ps.state = OP_MINUS_ERR_SPC
default:
goto parseErr
}
case OP_MINUS_ERR_SPC:
switch b {
case ' ', '\t':
continue
default:
nc.ps.state = MINUS_ERR_ARG
nc.ps.as = i
}
case MINUS_ERR_ARG:
switch b {
case '\r':
nc.ps.drop = 1
case '\n':
var arg []byte
if nc.ps.argBuf != nil {
arg = nc.ps.argBuf
nc.ps.argBuf = nil
} else {
arg = buf[nc.ps.as : i-nc.ps.drop]
}
nc.processErr(string(arg))
nc.ps.drop, nc.ps.as, nc.ps.state = 0, i+1, OP_START
default:
if nc.ps.argBuf != nil {
nc.ps.argBuf = append(nc.ps.argBuf, b)
}
}
case OP_P:
switch b {
case 'I', 'i':
nc.ps.state = OP_PI
case 'O', 'o':
nc.ps.state = OP_PO
default:
goto parseErr
}
case OP_PO:
switch b {
case 'N', 'n':
nc.ps.state = OP_PON
default:
goto parseErr
}
case OP_PON:
switch b {
case 'G', 'g':
nc.ps.state = OP_PONG
default:
goto parseErr
}
case OP_PONG:
switch b {
case '\n':
nc.processPong()
nc.ps.drop, nc.ps.state = 0, OP_START
}
case OP_PI:
switch b {
case 'N', 'n':
nc.ps.state = OP_PIN
default:
goto parseErr
}
case OP_PIN:
switch b {
case 'G', 'g':
nc.ps.state = OP_PING
default:
goto parseErr
}
case OP_PING:
switch b {
case '\n':
nc.processPing()
nc.ps.drop, nc.ps.state = 0, OP_START
}
case OP_I:
switch b {
case 'N', 'n':
nc.ps.state = OP_IN
default:
goto parseErr
}
case OP_IN:
switch b {
case 'F', 'f':
nc.ps.state = OP_INF
default:
goto parseErr
}
case OP_INF:
switch b {
case 'O', 'o':
nc.ps.state = OP_INFO
default:
goto parseErr
}
case OP_INFO:
switch b {
case ' ', '\t':
nc.ps.state = OP_INFO_SPC
default:
goto parseErr
}
case OP_INFO_SPC:
switch b {
case ' ', '\t':
continue
default:
nc.ps.state = INFO_ARG
nc.ps.as = i
}
case INFO_ARG:
switch b {
case '\r':
nc.ps.drop = 1
case '\n':
var arg []byte
if nc.ps.argBuf != nil {
arg = nc.ps.argBuf
nc.ps.argBuf = nil
} else {
arg = buf[nc.ps.as : i-nc.ps.drop]
}
nc.processAsyncInfo(arg)
nc.ps.drop, nc.ps.as, nc.ps.state = 0, i+1, OP_START
default:
if nc.ps.argBuf != nil {
nc.ps.argBuf = append(nc.ps.argBuf, b)
}
}
default:
goto parseErr
}
}
// Check for split buffer scenarios
if (nc.ps.state == MSG_ARG || nc.ps.state == MINUS_ERR_ARG || nc.ps.state == INFO_ARG) && nc.ps.argBuf == nil {
nc.ps.argBuf = nc.ps.scratch[:0]
nc.ps.argBuf = append(nc.ps.argBuf, buf[nc.ps.as:i-nc.ps.drop]...)
// FIXME, check max len
}
// Check for split msg
if nc.ps.state == MSG_PAYLOAD && nc.ps.msgBuf == nil {
// We need to clone the msgArg if it is still referencing the
// read buffer and we are not able to process the msg.
if nc.ps.argBuf == nil {
nc.cloneMsgArg()
}
// If we will overflow the scratch buffer, just create a
// new buffer to hold the split message.
if nc.ps.ma.size > cap(nc.ps.scratch)-len(nc.ps.argBuf) {
lrem := len(buf[nc.ps.as:])
nc.ps.msgBuf = make([]byte, lrem, nc.ps.ma.size)
copy(nc.ps.msgBuf, buf[nc.ps.as:])
nc.ps.msgCopied = true
} else {
nc.ps.msgBuf = nc.ps.scratch[len(nc.ps.argBuf):len(nc.ps.argBuf)]
nc.ps.msgBuf = append(nc.ps.msgBuf, (buf[nc.ps.as:])...)
}
}
return nil
parseErr:
return fmt.Errorf("nats: Parse Error [%d]: '%s'", nc.ps.state, buf[i:])
}
// cloneMsgArg is used when the split buffer scenario has the pubArg in the existing read buffer, but
// we need to hold onto it into the next read.
func (nc *Conn) cloneMsgArg() {
nc.ps.argBuf = nc.ps.scratch[:0]
nc.ps.argBuf = append(nc.ps.argBuf, nc.ps.ma.subject...)
nc.ps.argBuf = append(nc.ps.argBuf, nc.ps.ma.reply...)
nc.ps.ma.subject = nc.ps.argBuf[:len(nc.ps.ma.subject)]
if nc.ps.ma.reply != nil {
nc.ps.ma.reply = nc.ps.argBuf[len(nc.ps.ma.subject):]
}
}
const argsLenMax = 4
func (nc *Conn) processMsgArgs(arg []byte) error {
// Use separate function for header based messages.
if nc.ps.hdr >= 0 {
return nc.processHeaderMsgArgs(arg)
}
// Unroll splitArgs to avoid runtime/heap issues
a := [argsLenMax][]byte{}
args := a[:0]
start := -1
for i, b := range arg {
switch b {
case ' ', '\t', '\r', '\n':
if start >= 0 {
args = append(args, arg[start:i])
start = -1
}
default:
if start < 0 {
start = i
}
}
}
if start >= 0 {
args = append(args, arg[start:])
}
switch len(args) {
case 3:
nc.ps.ma.subject = args[0]
nc.ps.ma.sid = parseInt64(args[1])
nc.ps.ma.reply = nil
nc.ps.ma.size = int(parseInt64(args[2]))
case 4:
nc.ps.ma.subject = args[0]
nc.ps.ma.sid = parseInt64(args[1])
nc.ps.ma.reply = args[2]
nc.ps.ma.size = int(parseInt64(args[3]))
default:
return fmt.Errorf("nats: processMsgArgs Parse Error: '%s'", arg)
}
if nc.ps.ma.sid < 0 {
return fmt.Errorf("nats: processMsgArgs Bad or Missing Sid: '%s'", arg)
}
if nc.ps.ma.size < 0 {
return fmt.Errorf("nats: processMsgArgs Bad or Missing Size: '%s'", arg)
}
return nil
}
// processHeaderMsgArgs is for a header based message.
func (nc *Conn) processHeaderMsgArgs(arg []byte) error {
// Unroll splitArgs to avoid runtime/heap issues
a := [argsLenMax][]byte{}
args := a[:0]
start := -1
for i, b := range arg {
switch b {
case ' ', '\t', '\r', '\n':
if start >= 0 {
args = append(args, arg[start:i])
start = -1
}
default:
if start < 0 {
start = i
}
}
}
if start >= 0 {
args = append(args, arg[start:])
}
switch len(args) {
case 4:
nc.ps.ma.subject = args[0]
nc.ps.ma.sid = parseInt64(args[1])
nc.ps.ma.reply = nil
nc.ps.ma.hdr = int(parseInt64(args[2]))
nc.ps.ma.size = int(parseInt64(args[3]))
case 5:
nc.ps.ma.subject = args[0]
nc.ps.ma.sid = parseInt64(args[1])
nc.ps.ma.reply = args[2]
nc.ps.ma.hdr = int(parseInt64(args[3]))
nc.ps.ma.size = int(parseInt64(args[4]))
default:
return fmt.Errorf("nats: processHeaderMsgArgs Parse Error: '%s'", arg)
}
if nc.ps.ma.sid < 0 {
return fmt.Errorf("nats: processHeaderMsgArgs Bad or Missing Sid: '%s'", arg)
}
if nc.ps.ma.hdr < 0 || nc.ps.ma.hdr > nc.ps.ma.size {
return fmt.Errorf("nats: processHeaderMsgArgs Bad or Missing Header Size: '%s'", arg)
}
if nc.ps.ma.size < 0 {
return fmt.Errorf("nats: processHeaderMsgArgs Bad or Missing Size: '%s'", arg)
}
return nil
}
// ASCII numbers 0-9
const (
ascii_0 = 48
ascii_9 = 57
)
// parseInt64 expects decimal positive numbers. We
// return -1 to signal error
func parseInt64(d []byte) (n int64) {
if len(d) == 0 {
return -1
}
for _, dec := range d {
if dec < ascii_0 || dec > ascii_9 {
return -1
}
n = n*10 + (int64(dec) - ascii_0)
}
return n
}
|